Super Bowl LI (2017) prop betting
Quote: Wizard
After much thought and money lost, I have decided to retire from proposition bet making in the NFL. The primary reason for this is the game is played differently than it was in the early 2000's. The data on which my bets were made go back to 2000, when the rules were different. In an effort to make the game safer, the NFL has put a ball and chain on defenses, resulting in higher scoring games.
In conclusion, I am not saying that I will absolutely never make another prop bet. In fact, I made some on the AFC and NFC championship games, which went badly. I may make some small wagers that I have made every Super Bowl, just because old habits are hard to die, but I won't be betting thousands of dollars on each one like I used to do. More like in the low hundreds. Nor will I be promoting specific bets here or on the radio. Sorry, but for Super Bowl 2017, you're on your own.
Wizard,
I think your logic concerning not wanting to trust your model to make predictions about future Super Bowl prop bets has a lot of validity to it. But as someone who already has a trip to Vegas planned for Super Bowl weekend, I'm curious as to how valid that you feel your model is in relation to the rules for games from the era that the model was designed for.
Specifically, your model takes as an input the Over/Under for the games in computing Poisson Distributions. How confident are you the the model properly adjusted for high over/under games in that era (Rams, Saints, etc)? I would like to use your Super Bowl 50 model as one of my criteria for my own prop bets, adjusting for the effects of new rules (more accurate field goal kicking, more 2-pt conversions, etc.) but if you don't trust your model anymore for even those games, I would be less likely to trust it as well.
Quote: Ayecarumba
Yes. A safety in the course of an extra point kick attempt is now worth 1 point instead of the ball being dead as soon as the defense touched it. This rule was changed when the extra point attempt got moved further back a few seasons ago.
This was asked about in a former WoV trivia challenge. Then there was a huge fight here over the wording. Somewhere there is a thread about it.
Quote: tooncestdcI think your logic concerning not wanting to trust your model to make predictions about future Super Bowl prop bets has a lot of validity to it. But as someone who already has a trip to Vegas planned for Super Bowl weekend, I'm curious as to how valid that you feel your model is in relation to the rules for games from the era that the model was designed for.
Specifically, your model takes as an input the Over/Under for the games in computing Poisson Distributions. How confident are you the the model properly adjusted for high over/under games in that era (Rams, Saints, etc)? I would like to use your Super Bowl 50 model as one of my criteria for my own prop bets, adjusting for the effects of new rules (more accurate field goal kicking, more 2-pt conversions, etc.) but if you don't trust your model anymore for even those games, I would be less likely to trust it as well.
I'm sworn to secrecy about how my model works, although an astute mathematician could probably easily figure it out. I will say that every bet considers every game played from 2000 to 2014. It looks at how the spread and over/under affect every single prop. In other words, it tries to identify trends and then estimates how the spread and over/under for any game fits along those trends. I will say it does not directly use the Poisson distribution.
As you know, 58 is a very high over/under for the NFL. While you could put in an over/under of 5,800,000 and still get results, I would take anything it says for such a high over/under with more than just a grain of salt. I know this is all kind of vague but it is the best I can do.
Sorry if you came out here largely to bet props. If so, maybe I can meet you before or after the game for a free copy of my book.
There is also a log term affect, the more money taken out of the casino will affect future promotions.Quote: RSWell, they did disable the $5 VP shortly after you posting it on here (idk if posting here was the cause of it, IMO -- I think it's at least fair to say it very well could have been). I don't know of anyone from WOV that played it that wasn't already planning on playing it. The $5 VP being taken out not only hurt the APs financially (since we couldn't make money on it), but also hurt other non-APs who wouldn't been happy to sit on a machine for $10-20/hour leading up to the promo. Also wasted hours spent simming the game to come up with a bankroll required, other contingencies (what happens if they turn off the $5 machines mid-play after we're stuck $20k?), what kind of risk people want to take investing in the bankroll....then it gets muddied to all hell when the primary machines get taken out.
The owner said he missed the other plays. Had the owner knew about the other stuff prior to the promotion he would have excluded them or changed the pays. He didn't want to exclude or change things after it had started, he felt that would be bad for business.
We were lucky lots of people missed the other plays and no one posted up about them prior.
I don't think all the casinos are all reading the forums trying to get AP information. I do think there are darksiders that will use any and all AP information to their advantage. They will contact the casino(s) and tell them about their mistake, they are looking for a foot in the door. This is not speculation, it has happened. Let's not make their job easier by telling them where, what and when.
You can also find the Westgate and William at Full list of Super Bowl 51 prop bets from Westgate Superbook, William Hill.
Mohamed Sanu long reception 17.5 over +120, under -110 - put quite a bit more on the over. . .
Game not decided by 3 points -400 - in years past -500 has been a good bet, higher total makes this one even less likely to land on the three
Some other good ones I liked:
Matt Ryan to throw an interception +105; Brady to throw an interception before a TD +320; Ryan under 325.5 yards -110; Atlanta +7.5 -200; Atlanta -10.5 +525; under 51.5 +225; under four sacks -110; no Patriots TD in fourth quarter +220; no Patriots TD in second quarter +300; second quarter under 14.5 +115
Quote: TomGMatt Ryan to throw an interception +105
More than a couple max bets before the line moved, along with no interception +110 for a couple max bets
Not as good as a 50% edge last time the Patriots were in the Super Bowl, but nice to see it can still be done
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
The lines are the midpoint between the YES and No from the Wynn.
Quote: WizardI'd like to announce that DJTB has these bets against me:
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
The lines are the midpoint between the YES and No from the Wynn.
The mid-point isn't a fair line, BTW.
Quote: WizardI'd like to announce that DJTB has these bets against me:
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
The lines are the midpoint between the YES and No from the Wynn.
For some reason I like the scoreless quarter prop even though NE has scored in 31 consecutive quarters.
Quote: WizardI'd like to announce that DJTB has these bets against me:
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
The lines are the midpoint between the YES and No from the Wynn.
I assume it's nearly fair. I'm sure Mike can figure out the correct middle, but let's keep this simple.Quote: RSThe mid-point isn't a fair line, BTW.
Last year we had similar wagers and Mike was gonna give me the full value of his side. I said we're still friends. We can split the vig.
For the record, I may or may not have bet $100 on each. The amount is still private.
But I DID take the Yes on all four.
And if I only win one, it's fine if it's that one.
Quote: DJTeddyBearBy the way, I'm so clueless regarding sports betting that I actually had to ask, ok, since no money changes hands yet, if I put $100 on all the Yes bets, and only win the 2pt conv, then I'd owe $17.50, right?
$100 on each or total? Our lines or the Wynn lines?
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
If you risked $100 on each one on the YES side, at our odds, here is what would happen:
Safety: -100
Overtime: -100
score qtr: -100
2-pt conv: +35.40
Total = -264.60
Wouldn't $100 on the 2-pt conv be to WIN $282.50?Quote: Wizard$100 on each or total? Our lines or the Wynn lines?
Safety: +650
Overtime: +750
scoreless qtr: +525
2-pt conv.: +282.5
If you risked $100 on each one on the YES side, at our odds, here is what would happen:
Safety: -100
Overtime: -100
score qtr: -100
2-pt conv: +35.40
Total = -264.60
Thus...
Safety: -100
Overtime: -100
scote qrt: -100
2-pt conv: +282.50
Total = -$17.50?
Not lookin' at actual lines, but DJTB probably got some +EV bets in.
Quote: lilredroosterIt would be interesting to hear how some of the posters would evaluate a prop bet. i.e.: if the prop bet was will either team run a kick back for a TD. Would you look at how often this was done in a Super Bowl? Or how often it was done during the full season? Or all seasons the last 10 years? Or how often each team did it during the season? If it was me I think I would do it by comparing each kick returner's total yards per run compared to others for the same season and how many touchdowns did each returner score in the past season including playoff games on kick returns. And how many TDs did each team allow on kick returns. I think I would give equal weight to each stat. And calculate a % based on that. But there might be better ways to do it. Does it make a difference that it's a Super Bowl and not an ordinary game. I don't really know. But I would guess no.
It may very well be 'stadium' dependent. Kicking into the wind there will be far fewer touchbacks. Indoor stadium changes that. Indoor stadium less actual runbacks. No actual evidence analysis by me. I would bet against KO return in indoor game, on in outdoor game.
Quote: RomesWouldn't $100 on the 2-pt conv be to WIN $282.50?
Thus...
Safety: -100
Overtime: -100
scote qrt: -100
2-pt conv: +282.50
Total = -$17.50?
You're right. I misread his request to want to know what would happen if he had my side.
You know who you are. ;)
Ryan to throw an interception -115; Brady to throw an interception +160; no defensive/special teams TD -185. The percentages implied by those odds cannot all be true (I point this one out, because all three are good bets)
Also Ryan to throw an interception -115; Devin McCourtney no interception -500; Malcom Butler no interception -375 (again all three are good bets, and there is no way McCourtney has a 20% chance of interception if Ryan only has a 50% of throwing one)
Quote: SOOPOOIt may very well be 'stadium' dependent. Kicking into the wind there will be far fewer touchbacks. Indoor stadium changes that. Indoor stadium less actual runbacks. No actual evidence analysis by me. I would bet against KO return in indoor game, on in outdoor game.
Rather than the player returning the kick, I think the coach has the biggest influence on the return. Some coaches instruct their players to always take the knee.
The other factor is the kicker. Does he have the leg to clear the end zone everytime?
Ended up pretty heavy on Atlant +3 -105; bought off a lot on Patriots +7.5 -360. Another great bet.
If not for having a lot on under 60 and 59.5, would have a profit virtually locked in by kickoff.
Quote: TomGFalcons 2q +0.5 -105. Great bet.
Nice! Stations has that at -130.
I show the second quarter is tied 14.36% of the time. If we assume the Falcons have a 47% chance to win it if not tied, then your chances of winning are 47% + 14.36% = 61.36%. Laying 105, you would have a 19.8% advantage.
If you want more Atlanta +3, I'll give you even money, as I like the other side.
Here is my count of games by over/under and how many times there were under 7.5 punts as well as the percentage under.
| Over/Under | Games | Under 7.5 Punts | Ratio under 7.5 punts |
|---|---|---|---|
| 30.5 | 10 | 0 | 0.00% |
| 30.5 | 10 | 0 | 0.00% |
| 31 | 6 | 0 | 0.00% |
| 31.5 | 14 | 0 | 0.00% |
| 32 | 14 | 0 | 0.00% |
| 32.5 | 22 | 2 | 9.09% |
| 33 | 78 | 14 | 17.95% |
| 33.5 | 58 | 0 | 0.00% |
| 34 | 108 | 10 | 9.26% |
| 34.5 | 110 | 12 | 10.91% |
| 35 | 98 | 14 | 14.29% |
| 35.5 | 112 | 8 | 7.14% |
| 36 | 162 | 24 | 14.81% |
| 36.5 | 168 | 30 | 17.86% |
| 37 | 310 | 52 | 16.77% |
| 37.5 | 326 | 58 | 17.79% |
| 38 | 218 | 52 | 23.85% |
| 38.5 | 192 | 32 | 16.67% |
| 39 | 220 | 40 | 18.18% |
| 39.5 | 202 | 32 | 15.84% |
| 40 | 220 | 48 | 21.82% |
| 40.5 | 282 | 66 | 23.40% |
| 41 | 344 | 66 | 19.19% |
| 41.5 | 290 | 66 | 22.76% |
| 42 | 302 | 92 | 30.46% |
| 42.5 | 274 | 84 | 30.66% |
| 43 | 348 | 88 | 25.29% |
| 43.5 | 360 | 114 | 31.67% |
| 44 | 358 | 118 | 32.96% |
| 44.5 | 304 | 92 | 30.26% |
| 45 | 302 | 120 | 39.74% |
| 45.5 | 208 | 68 | 32.69% |
| 46 | 224 | 82 | 36.61% |
| 46.5 | 202 | 64 | 31.68% |
| 47 | 230 | 84 | 36.52% |
| 47.5 | 226 | 80 | 35.40% |
| 48 | 176 | 70 | 39.77% |
| 48.5 | 128 | 40 | 31.25% |
| 49 | 92 | 40 | 43.48% |
| 49.5 | 78 | 32 | 41.03% |
| 50 | 78 | 34 | 43.59% |
| 50.5 | 84 | 42 | 50.00% |
| 51 | 82 | 42 | 51.22% |
| 51.5 | 52 | 28 | 53.85% |
| 52 | 42 | 22 | 52.38% |
| 52.5 | 28 | 18 | 64.29% |
| 53 | 22 | 10 | 45.45% |
| 53.5 | 40 | 16 | 40.00% |
| 54 | 30 | 10 | 33.33% |
| 54.5 | 24 | 8 | 33.33% |
| 55 | 20 | 8 | 40.00% |
| 55.5 | 22 | 10 | 45.45% |
| 56 | 16 | 8 | 50.00% |
| 56.5 | 14 | 12 | 85.71% |
| 57 | 14 | 10 | 71.43% |
| 57.5 | 6 | 2 | 33.33% |
| 58 | 6 | 6 | 100.00% |
| 58.5 | 0 | 0 | #DIV/0! |
| 59 | 4 | 2 | 50.00% |
| 59.5 | 6 | 6 | 100.00% |
I know there isn't a lot of data for spreads of 58 to 59, like we have for this Super Bowl, but if you draw a linear regression line through this data, and eyeballing it the relationship looks linear, you would get a probability of under 7.5 punts of 62.465%, assuming an over/under of 58.5. That equates to a fair line of -166. I just got -145 at 5dimes. That is the same line they have at Wynn. Stations is -155.
Quote: RShttp://sportsbettingsites.org/betting-tools/no-vig-calculator/
Not lookin' at actual lines, but DJTB probably got some +EV bets in.
Quote: beachbumbabsAhem.
You know who you are. ;)
I finally had a chance to look at this. Note that I'm still uncertain about it.
Is Babs agreeing with RS that I'm getting better than fair payout, or is Babs suggesting that Mike has done me dirty? I suspect the former.
Frankly, I don't care either way. All I know is, when Mike and I originally started our friendly SB bets, he wanted to give me the full -x value. While I knew that splitting the vig down the middle wasn't correct, I knew that the middle was at least fairer than what he wanted to do. After all, if I won, I would get more than what I'd have gotten if I went to a sports book, while Mike would pay less than going to a book.
Bottom line, thanks for the link to that tool, but please stop blaming Mike for not suggesting doing that math. Truth is, I had to ask him twice not to give me the full sports book -x value.
After all, it was supposed to be friendly bets.
Quote: DJTeddyBearQuote: beachbumbabsAhem.
You know who you are. ;)
I finally had a chance to look at this. Note that I'm still uncertain about it.
Is Babs agreeing with RS that I'm getting better than fair payout, or is Babs suggesting that Mike has done me dirty? I suspect the former.
Frankly, I don't care either way. All I know is, when Mike and I originally started our friendly SB bets, he wanted to give me the full -x value. While I knew that splitting the vig down the middle wasn't correct, I knew that the middle was at least fairer than what he wanted to do. After all, if I won, I would get more than what I'd have gotten if I went to a sports book, while Mike would pay less than going to a book.
Bottom line, thanks for the link to that tool, but please stop blaming Mike for not suggesting doing that math. Truth is, I had to ask him twice not to give me the full sports book -x value.
After all, it was supposed to be friendly bets.
I think babs put in some different bet, or talking about something else, as she entered arbitragable lines.
DeMar DeRozan points vs. Julio Jones Longest reception.
DD's per game average and the number for JJ's long td are just about the same. But someone let me know DD is hurt.
Dustin Brown Pts vs Brady INTs.
I only ran through it once, so forgive an error. But I got DB scoring in about 27% of games. He's playing the Capitols, the top defense.
Brady throwing an INT is only +130.
It's not usually rocket surgery, especially when I do it.
Edelman total in mid 90s. FSU are 7 point favs, game total 150.5.
Quote: WizardI know I said I wasn't going to recommend props this year but I'm going to give you just one -- Under 7.5 punts.
4 of them in the 1st alone.
Quote: DJTeddyBearQuote: beachbumbabsAhem.
You know who you are. ;)
I finally had a chance to look at this. Note that I'm still uncertain about it.
Is Babs agreeing with RS that I'm getting better than fair payout, or is Babs suggesting that Mike has done me dirty? I suspect the former.
Frankly, I don't care either way. All I know is, when Mike and I originally started our friendly SB bets, he wanted to give me the full -x value. While I knew that splitting the vig down the middle wasn't correct, I knew that the middle was at least fairer than what he wanted to do. After all, if I won, I would get more than what I'd have gotten if I went to a sports book, while Mike would pay less than going to a book.
Bottom line, thanks for the link to that tool, but please stop blaming Mike for not suggesting doing that math. Truth is, I had to ask him twice not to give me the full sports book -x value.
After all, it was supposed to be friendly bets.
No, no, no, no, no. This wasn't about you at all. Sorry for anything I did that made you think that. Geez.
To get back $100 you have to bet
(i) $87.50 to win $12.50
(ii) $16.67 to win $83.33
Thus the total bet is $104.17 (also known as Book Value).
This means fair chances are 87.50/104.17 and 16.67/104.17 which are 84% and 16% (those used to first principles will notice the bets to a 48 book are 42 to win 6 and 8 to win 40, so 50 to win 48.)
Thus true odds are 16 to 84 which simplified are 4/21 and 21/4 or 1 to 5.25 and 5.25 to 1.
Summary
Take each possibility and work out the bet size required to get back $100.
Add up the total money required to cover all bets - for a bookmaker this will be more than $100.
Take each possibility and divide the bet by a factor of "Book Value"/100.
This gives the fair bet size required, from which the odds can be calculated.
I thought the drones they used were wild.
Quote: GWAEHope no one took lady gaga to show cleavage.
I thought the drones they used were wild.
I changed the channel. Would rather commit suicide than watch her/him.
Quote: Ibeatyouraces4 of them in the 1st alone.
Up to 6 punts.
Quote: IbeatyouracesI changed the channel. Would rather commit suicide than watch her/him.
I actually like her songs but I also like top 40.
I'm on Margaritas rather than pints!Quote: IbeatyouracesUp to 6 punts.
Quote: GWAEHope no one took lady gaga to show cleavage.
I thought the drones they used were wild.
What did they do with the drones?
Quote: AyecarumbaWhat did they do with the drones?
She was on the roof singing aND they have about 200 drones with different color lights behind her in the sky making designs.
Quote: GWAEI actually like her songs but I also like top 40.
The jury is still out on the bolded part.
Quote: GWAEI actually like her songs but I also like top 40.
I like neither, but couldn't help but be impressed with her live skills.
Quote: Mission146I like neither, but couldn't help but be impressed with her live skills.
That is the first thing I look for in all performances. I hate lip syncing. She was definitely signing and playing the piano.
Quote: GWAEShe was on the roof singing aND they have about 200 drones with different color lights behind her in the sky making designs.
I thought it was just laser projections on the roof. The lights weren't very bright. I would have appreciated it more if the they showed a close up of the drones.
Quote: GWAEThat is the first thing I look for in all performances. I hate lip syncing. She was definitely signing and playing the piano.
I think she was lip-synching, though augmenting with a live mike, until she sat down at the piano. That was live. Then I think it went back to lip-synching at the very end.
I did not understand how they did the lighting behind her at the start. I'm glad you guys figured out that was drones.
I don't understand still how they could control a couple thousand devices individually. There aren't that many available frequencies. Maybe they are addressable, like cell phones, a couple times a second.
Things have come a long, long way since I was a theatrical lighting designer. I stopped doing that in early '85. It's a whole new world, though they still use most/all of what we had then. Amazing field.
